Topic: 0 byte file -- not being able to write -- same issue?

[raddygast]

Today (for the second time I was using the MT for a mission-critical purpose), it failed.

Firmware: 1.2.0  (I hadn't checked for a newer one, d'oh)
1/4" TRS inputs (Left and Right) with +27dB boost on, M setting, de-linked levels, 16 bit 44.1 KHz

Kingston Elite Pro 50x GF 1GB 1024-S (supported officially by m-audio) ... which was formatted on the MT under 1.2.0 (not sure if it came formatted, but was constantly reformatted in the first days of me fooling around with 1.2.0)

Now, I've always been very antsy about the way this thing records.... That is, you can record over an hour and then when you stop the recording you have to wait (and sweat bullets) while it "writes file." If that part fails, the data is likely lost. Is it just me or does this feel really scary?

Anyway, today I noticed about half an hour into recording, the recording was stopped. It was on track 1/1  00:00:00/00:00:00

I was perplexed, because I had about an hour left. So I hit the record button to stop the file... thinking the numbers were just strange. Then the write file window came up (Writing File...) but it just stayed on screen forever. I tried powering off and a popup said you can't power off while still recording. And since the file wasn't being written successfully, it wouldn't let me power off.

When I got home, it was finally off (I did have auto-off at 15 min, maybe that's it)... when I powered on, it showed there was a file there ... but it was 0 byte. It was never written successfully, so I don't know how the file even got there. I thought if the MT doesn't finalize a file, it's just not there.

Anyway, before I had a chance to peruse these threads, I had already reformatted the CF card, so I don't even get to try the method using chkdsk /f on the MT, then recovery of the WAV file through CEP (what is CEP by the way?)


Now, what could cause my initial problem? What happens when the MT can't finish writing a file? Is there a way to hard-reset the unit? Should you eject the card manually or not?

I hope (hope hope hope) that the bug I got (suddenly the counters read 00:00:00/00:00:00 which I thought was suspicious, and the recording stopped, and it couldn't record any more) is related to the incorrect time remaining counters that were just fixed. But maybe not.

maybe this problem is going to happen again and again. I really lost face today when I had to admit to my client that the whole interview may have been lost.

Anyone else have this issue or know anything about it?

[raddygast]

Well, with firmware 1.2.3 and a freshly formatted (in the MT) CF card, I tried recording nothing (i.e. didn't plug the mics in). Tried a bunch of recordings, between a few minutes and half an hour, no problem. So I let a really long one go... last time I checked it was at 1h29m and still going (9m remaining).

Next time I checked the recording was stopped with the dreaded 00:00:00/00:00:00.

I think this was the exact prob I had earlier today (except it happened long before the disk was actually full!) The first time I think I noticed it was stopped and tried to hit record to re-start, and only then did the "Writing!" dialog come up (and freeze forever).

This time (in testing) I knew the CF was full. I didn't try hitting the record button again. I checked time remaining in Record Settings and it was indeed 00:00:00.

In Windows the drive showed up as pretty much full except for a few KB, even though there was a single 0 byte file on it.

I tried properties on the drive, tools tab, error-checking, ticked repair filesystem errors. Then suddenly the wav file was 1GB in size but corrupt. So I renamed it to raw and opened in Adobe Audition with some guessing, and it seems ok (dunno because the mics weren't hooked up, there's a little intermittent pulsing noise which I'm wondering may be internal noise due to the 27dB boost and my maxed out levels on the left channel).

So... what the hell can be causing this? I have a 1GB CF card only. Shouldn't be an issue -- shouldn't it gracefully write the file? I should mention it was not connected to power during this whole ordeal -- running off bat only. But the bat didn't run out.

I am very perplexed. And annoyed as hell, because it happened with 1.2.3 so there's no chance the fw upgrade fixed it.

[branas]

I would try it again but record "something" so you know what is really happening. Also, I would try stopping it just short of 1m or two remaining, so it has room to buffer and write the end of the file, this may be the key. I know, they should think of this when calculating the remaining time, but, alas. I would hope the noise you refer to is not internal noise as it is rated at .00265% THD.

I had no problems recording 3.5 hours on a 4GB Hitachi microdrive the other night, first file was 1:45:00 or so, second file was around 1:30:00, recorded at 24bit/44.1, running 1.2.3 firmware.

[smokydays]

Anyway, before I had a chance to peruse these threads, I had already reformatted the CF card, so I don't even get to try the method using chkdsk /f on the MT, then recovery of the WAV file through CEP (what is CEP by the way?)

CEP = Cool Edit Pro.  Because of a buyout, it is now Adobe Audition.

[raddygast]

I did the test with the supplied 64MB CF card. Unfortunately it was just fine with that card (which can only hold 4.5 mins or so) so I don't know whether it's my CF card, or some aspect of a longer recording (memory or buffer related) or something which just gets aggravated the longer a track has been recording.

I tried it again with my card, and here are the exact findings in case anyone else wants to compare:

This is exactly what happens. I record one long file (after formatting the CF card on the MT). As soon as I start recording, I hit rec-pause, then start again (a couple of times, to make sure it's working). But I leave the same file going. I let it record on its own till the end.

When I come back to check it, I notice it's in stop mode, and reads "file 1/1 00:00:00/00:00:00"

I hit the rec button again, and it tries to start up a new file with a higher number, but after a second it just drops out of that. At this point (I believe, from memory) there is no "file 1/1" or "file 1/2" at the bottom.... it's just blank, as if the card is blank. However, rec time remaining remains at 00:00:00.

If I hit the rec button again, the window pops up that says "Writing file...." which stays up forever. If i try to power off I get a warning saying "The recording must be stopped before shutting down."  The only thing I can do is hit delete, which asks me "hit delete again to confirm" and when I hit delete again, the "Writing file...." screen is replaced by "Deleting file....." but it also does not complete.

The menu button does enter the menu (record settings, record time available = 00:00:00) but once you back out of the menu it's still stuck on "Writing file...." or "Deleting file..."

I have tested this twice now with my 1GB Kingston Elite Pro 50x CF card (1024-S) and it happened both times (if I let it run to the end). I tried it once with the supplied Kodak 64MB card that came with the MT, and that one saved the file gracefully at the end of the 4 minutes capacity. It may be related to large capacity or full buffers or something, I'm not sure. Could it be my CF card? It's supported officially.

I will try formatting my CF card through XP as per m-audio's instructions to see if this still happens.

[raddygast]

I tried formatting the CF card under Windows XP, but I still get the exact same replicable problem.

Tried a 256MB Verbatim CF card (from my camera) and the problem is almost identical, except the "media full" popup occurs a bit before the end of the media. With 256MB you get 24:18 recording time, and with about 2:45 left or so it craps out. Only on this card it gets hopelessly corrupted such that chkdsk /f can't even fix it.

I really don't think it's my CF card. I am very wary of this MT unit.... I wonder if mine's defective or if this is a widespread problem. Anyone try recording to the end of your media?

[branas]

I don't want to belittle the issue you are having, but I am wondering, (I am new so this may be stupid), how often really will you let it record all the way to the end of the media? I would not think that it would be normal practice to do so.

[raddygast]

Good point, and an update:

If I stop the recording like 50 secs before the media is full, it saves correctly. Then if I start a new file, let that record until media full, it saves that gracefully too. Problem seems to occur (on my 1GB) card when I let one file span the whole space fully. On the 256MB card, it happens earlier.

Your point is well-taken. However, this whole testing process was started because my recorder crapped out on my *in the field* while I was recording an important interview. It was not out of space -- it just stopped about 30-40 minutes into a file. I can't figure out how to duplicate that, but the symptoms it showed then were exactly the same as the symptoms I get when I run a recording all the way to the end: i.e. the file shows as 1/1 00:00:00/00:00:00 and is 0 bytes in length, and trying to start another file results in the "Writing file..." message forever.

M-Audio are going to exchange my unit, but I have a feeling it's not defective and the card is ok too. Maybe it's a firmware bug.

[smokydays]

Not that I have a microtracker, but i have read alot of threads about the benefits of updating the firmware and issues that it has solved.  I would do that first thing when i got the new one.

[raddygast]

Have the newest fw on mine, and I already plan to update the fw on the replacement unit as soon as I get it home.

[jtessier]

Have the newest fw on mine, and I already plan to update the fw on the replacement unit as soon as I get it home.

Good news and bad news,

Bad news is that on a 1GB card I can get the problem to happen for me too.  Good news is that M-Audio has been putting out fairly regular firmware updates.  Also I do not see this problem on larger media so far as I can tell. If it hits the 2GB per file size limit, then it successfully saves. I guess if you could afford it, a 4GB CF or Microdrive would be a intermediate fix until the firmware is corrected. Lastly, , like you, I don't see the problem on smaller media. The 64 that comes with it and a 256 I have both work fine all the way to the end.

J.T.

[raddygast]

What brand of card do you use jtessier?

[raddygast]

Replaced the unit for a brand new one, updated its firmware, and I can confirm the problem exists in exactly the same way.

To top it all off, the nav button on the new unit is a bit crappy. When I first got it it would get stuck in the "up" direction and not spring back. I've jiggled the crap out of it and now it's much better but still "sluggish" and you can feel it scraping slightly when you push up. Anyone know if this is easily fixable by myself?

[jtessier]

What brand of card do you use jtessier?

Sorry, been busy. It was a 1GB Lexar 80x card. I did have time to try again a couple times and found that sometimes it did work.  I don't have much else to add right now.  I'm going to try and see if there's some particular sequence of steps that I followed that either causes the problem or prevents the problem.

J.T.

[raddygast]

Thanks for the update. Strange that the problem should be variable. I have tried it over thirty times and each time, the same problem occurs at the end of the procedure.

I have a theory that it has to do with the remaining time calculation being off by a few seconds or so. The MT thinks it has, say, five seconds remaining (plus a little leeway to flush the buffer and close the file), but maybe in reality it has really run out of time. So the firmware is unprepared when it seems to have run out, and the file isn't properly closed.

I tried closing a file with 7 seconds remaining, and it was fine. Then I recorded another file, and it quit after 1 second (even though it said 7 secs remaining) -- saying "Media is full!"  So I do believe it is something to do with improper calculations.